https://aiweekly.co/alerts/df26-benchmark-humans-and-detectors-near-chance-on-ai-video AI Weekly en 2026-09-10T15:57:50+00:00 DF26 benchmark: humans and detectors near chance on AI video https://aiweekly.co/alerts/dcp-audit-logs-zero-recoveries-across-96-ai-agent-episodes AI Weekly en 2026-09-10T15:57:50+00:00 DCP audit logs zero recoveries across 96 AI-agent episodes https://aiweekly.co/alerts/wearableqa-benchmark-stretches-14-llms-from-196-to-729 AI Weekly en 2026-09-10T15:57:50+00:00 WearableQA benchmark stretches 14 LLMs from 19.6% to 72.9% https://aiweekly.co/alerts/nvidia-signs-eight-australian-operators-for-2gw-ai-buildout-by-2027 AI Weekly en 2026-09-10T15:29:09+00:00 Nvidia Signs Eight Australian Operators for 2GW AI Buildout by 2027 https://aiweekly.co/alerts/router-prior-bias-preserves-moe-routing-during-post-training AI Weekly en 2026-09-10T15:12:01+00:00 Router Prior Bias preserves MoE routing during post-training https://aiweekly.co/alerts/lambert-expect-lossy-self-improvement-not-fast-takeoff AI Weekly en 2026-09-10T15:12:01+00:00 Lambert: expect 'lossy self-improvement,' not fast takeoff https://aiweekly.co/alerts/bengio-warns-recent-ai-lab-tests-preview-losing-control AI Weekly en 2026-09-10T12:12:01+00:00 Bengio warns recent AI lab tests preview losing control https://aiweekly.co/alerts/anthropic-alignment-lead-10-chance-ai-kills-all-humans AI Weekly en 2026-09-10T12:12:01+00:00 Anthropic alignment lead: >10% chance AI kills all humans https://aiweekly.co/alerts/moonshot-ai-explores-dual-hong-kong-shanghai-ipo-at-50b AI Weekly en 2026-09-10T10:55:46+00:00 Moonshot AI Explores Dual Hong Kong, Shanghai IPO at ~$50B https://aiweekly.co/alerts/bench-tests-frontier-llms-on-their-own-infrastructure-stack AI Weekly en 2026-09-10T09:30:45+00:00 Φ-Bench Tests Frontier LLMs on Their Own Infrastructure Stack https://aiweekly.co/alerts/deepseek-posts-v41-flash-552b-moe-8b-active-1m-context AI Weekly en 2026-09-10T09:12:01+00:00 DeepSeek posts V4.1-Flash: 552B MoE, 8B active, 1M context https://aiweekly.co/alerts/apple-watch-series-12-gets-ai-to-summarize-conversations AI Weekly en 2026-09-10T09:12:01+00:00 Apple Watch Series 12 gets AI to summarize conversations https://aiweekly.co/alerts/llms-design-inventory-policies-cut-costs-up-to-30-in-study AI Weekly en 2026-09-10T09:12:01+00:00 LLMs Design Inventory Policies, Cut Costs Up to 30% in Study https://aiweekly.co/alerts/aiide-2026-opens-registration-for-first-brazil-edition AI Weekly en 2026-09-10T09:12:01+00:00 AIIDE 2026 opens registration for first Brazil edition https://aiweekly.co/alerts/zoox-woos-sf-riders-with-influencers-as-waymo-holds-lead AI Weekly en 2026-09-10T07:59:30+00:00 Zoox Woos SF Riders With Influencers as Waymo Holds Lead https://aiweekly.co/alerts/alibaba-bytedance-pay-chinese-pros-15-74-to-train-ai AI Weekly en 2026-09-10T06:43:07+00:00 Alibaba, ByteDance Pay Chinese Pros $15-$74 to Train AI https://aiweekly.co/alerts/bench-claude-opus-5-tops-3653-on-llm-infrastructure-tasks AI Weekly en 2026-09-10T03:54:22+00:00 Φ-Bench: Claude Opus 5 Tops 36.53% on LLM Infrastructure Tasks https://aiweekly.co/alerts/vdiff-bench-grok-43-trails-kimi-on-low-level-image-diffs AI Weekly en 2026-09-10T03:54:22+00:00 VDiff-Bench: Grok 4.3 trails Kimi on low-level image diffs https://aiweekly.co/alerts/a-thought-v2-halves-reasoning-length-via-latent-compression AI Weekly en 2026-09-10T03:54:22+00:00 A*-Thought-V2 Halves Reasoning Length via Latent Compression https://aiweekly.co/alerts/show-harness-lets-frontier-vlms-drive-robots-zero-shot AI Weekly en 2026-09-10T03:27:32+00:00 Show-Harness Lets Frontier VLMs Drive Robots Zero-Shot https://aiweekly.co/alerts/programmable-world-model-decouples-state-from-video-renderer AI Weekly en 2026-09-10T03:27:32+00:00 Programmable World Model Decouples State From Video Renderer https://aiweekly.co/alerts/anthropic-researcher-coxon-quits-calls-ai-race-endgame AI Weekly en 2026-09-10T03:12:01+00:00 Anthropic Researcher Coxon Quits, Calls AI Race 'Endgame' https://aiweekly.co/alerts/anthropic-pledges-350m-to-prep-policy-for-ai-labor-shock AI Weekly en 2026-09-10T03:12:01+00:00 Anthropic pledges $350M to prep policy for AI labor shock https://aiweekly.co/alerts/study-maximize-profit-prompt-makes-llms-downplay-risks AI Weekly en 2026-09-10T03:12:01+00:00 Study: 'maximize profit' prompt makes LLMs downplay risks https://aiweekly.co/alerts/hugging-face-publishes-flagship-so-101-arm-build-guide AI Weekly en 2026-09-10T03:12:01+00:00 Hugging Face publishes flagship SO-101 arm build guide https://aiweekly.co/alerts/q2d-web-pits-13-retrievers-on-190m-doc-agentic-rag-corpus AI Weekly en 2026-09-10T03:12:01+00:00 Q2D-Web pits 13 retrievers on 190M-doc agentic RAG corpus https://aiweekly.co/alerts/google-deepmind-partners-with-cfs-to-steer-sparc-fusion-plasma AI Weekly en 2026-09-10T03:12:01+00:00 Google DeepMind partners with CFS to steer SPARC fusion plasma https://aiweekly.co/alerts/openai-agents-reached-10-undisclosed-sites-six-groups-say AI Weekly en 2026-09-10T01:59:03+00:00 OpenAI Agents Reached 10+ Undisclosed Sites, Six Groups Say https://aiweekly.co/alerts/meta-seizes-muse-handles-from-british-rock-band-for-ai-agent AI Weekly en 2026-09-10T01:59:03+00:00 Meta Seizes @Muse Handles From British Rock Band for AI Agent https://aiweekly.co/alerts/apples-siri-ai-ships-sept-14-on-ios-27-trained-on-gemini AI Weekly en 2026-09-10T00:27:39+00:00 Apple's Siri AI Ships Sept 14 on iOS 27, Trained on Gemini https://aiweekly.co/alerts/ninth-circuit-lifts-cfaa-injunction-on-perplexity-comet-browser AI Weekly en 2026-09-10T00:12:01+00:00 Ninth Circuit lifts CFAA injunction on Perplexity Comet browser https://aiweekly.co/alerts/apple-defends-new-apple-watchs-always-listening-features AI Weekly en 2026-09-10T00:12:01+00:00 Apple defends new Apple Watch's always-listening features https://aiweekly.co/alerts/microsoft-edge-team-ai-code-surge-slows-extension-reviews AI Weekly en 2026-09-09T22:58:09+00:00 Microsoft Edge Team: AI Code Surge Slows Extension Reviews https://aiweekly.co/alerts/adobe-embeds-firefly-veo-runway-kling-luma-in-premiere AI Weekly en 2026-09-09T21:30:52+00:00 Adobe Embeds Firefly, Veo, Runway, Kling, Luma in Premiere https://aiweekly.co/alerts/anthropic-claude-mythos-5-uploaded-malicious-pypi-packages AI Weekly en 2026-09-09T21:12:01+00:00 Anthropic: Claude Mythos 5 uploaded malicious PyPI packages https://aiweekly.co/alerts/anthropic-details-four-claude-cyber-incidents-metr-to-audit AI Weekly en 2026-09-09T19:58:07+00:00 Anthropic Details Four Claude Cyber Incidents, METR to Audit https://aiweekly.co/alerts/popken-and-raman-56-experts-across-24-countries-on-llm-use AI Weekly en 2026-09-09T18:12:01+00:00 Popken and Raman: 56 experts across 24 countries on LLM use https://aiweekly.co/alerts/pangram-and-grammarlys-authorship-fuel-a-campus-ai-arms-race AI Weekly en 2026-09-09T18:12:01+00:00 Pangram and Grammarly's Authorship fuel a campus AI arms race https://aiweekly.co/alerts/vehtari-team-benchmarks-11-shortcuts-for-bayesian-logo-cv AI Weekly en 2026-09-09T18:12:01+00:00 Vehtari team benchmarks 11 shortcuts for Bayesian LOGO-CV https://aiweekly.co/alerts/unitedhealth-sued-over-nh-predict-ai-in-medicare-denials AI Weekly en 2026-09-09T18:12:01+00:00 UnitedHealth sued over nH Predict AI in Medicare denials https://aiweekly.co/alerts/anthropic-model-sees-labor-share-at-452-in-extreme-ai-case AI Weekly en 2026-09-09T16:59:32+00:00 Anthropic Model Sees Labor Share at 45.2% in Extreme AI Case https://aiweekly.co/alerts/jd-cloud-moore-threads-plan-100000-gpu-chinese-silicon-cluster AI Weekly en 2026-09-09T16:59:32+00:00 JD Cloud, Moore Threads Plan 100,000-GPU Chinese-Silicon Cluster https://aiweekly.co/alerts/deepseek-harness-bug-let-agents-kill-sandbox-in-one-command AI Weekly en 2026-09-09T16:59:32+00:00 DeepSeek Harness Bug Let Agents Kill Sandbox in One Command https://aiweekly.co/alerts/llms-recognize-impossible-questions-but-answer-them-anyway AI Weekly en 2026-09-09T15:54:09+00:00 LLMs Recognize Impossible Questions but Answer Them Anyway https://aiweekly.co/alerts/transformers-encode-partner-expertise-early-act-on-it-late AI Weekly en 2026-09-09T15:54:09+00:00 Transformers encode partner expertise early, act on it late https://aiweekly.co/alerts/self-listening-paper-targets-voice-ai-anchor-interruption AI Weekly en 2026-09-09T15:54:09+00:00 'Self-Listening' Paper Targets Voice AI Anchor Interruption https://aiweekly.co/alerts/synthgait-19k-19272-synthetic-clips-train-monocular-gait-models AI Weekly en 2026-09-09T15:54:09+00:00 SynthGait-19K: 19,272 synthetic clips train monocular gait models https://aiweekly.co/alerts/mothers-openai-wrongful-death-suit-cites-289-page-4o-chat AI Weekly en 2026-09-09T15:28:12+00:00 Mother's OpenAI Wrongful-Death Suit Cites 289-Page 4o Chat https://aiweekly.co/alerts/man-sues-openai-altman-over-chatgpt-jesus-delusion-chats AI Weekly en 2026-09-09T15:12:01+00:00 Man sues OpenAI, Altman over ChatGPT Jesus-delusion chats https://aiweekly.co/alerts/lambert-ai-is-still-a-rounding-error-in-everyday-life AI Weekly en 2026-09-09T15:12:01+00:00 Lambert: AI is still 'a rounding error' in everyday life https://aiweekly.co/alerts/analog-devices-to-acquire-alif-semiconductor-for-135b-cash AI Weekly en 2026-09-09T13:56:10+00:00 Analog Devices to Acquire Alif Semiconductor for $1.35B Cash https://aiweekly.co/alerts/pachocki-says-no-lab-has-solved-ai-alignment-urges-slowdowns AI Weekly en 2026-09-09T12:41:26+00:00 Pachocki says no lab has solved AI alignment, urges slowdowns https://aiweekly.co/alerts/harvey-raises-550m-at-156b-co-led-by-lightspeed-and-diffusion AI Weekly en 2026-09-09T12:30:49+00:00 Harvey Raises $550M at $15.6B, Co-Led by Lightspeed and Diffusion https://aiweekly.co/alerts/european-commissions-ai-and-culture-consultation-closes-sept-11 AI Weekly en 2026-09-09T12:12:01+00:00 European Commission's AI-and-culture consultation closes Sept 11 https://aiweekly.co/alerts/rettberg-diagnoses-llm-misaligned-citations-as-fabrication AI Weekly en 2026-09-09T12:12:01+00:00 Rettberg diagnoses LLM 'misaligned citations' as fabrication https://aiweekly.co/alerts/berkeleydeepmind-paper-llms-shift-essay-meaning-even-on-grammar-edits AI Weekly en 2026-09-09T12:12:01+00:00 Berkeley/DeepMind paper: LLMs shift essay meaning, even on grammar edits https://aiweekly.co/alerts/marigold-v2-cuts-kitti-eth3d-absrel-16-26-with-qlora-tuned-dit AI Weekly en 2026-09-09T10:56:18+00:00 Marigold V2 Cuts KITTI, ETH3D AbsRel 16-26% With QLoRA-Tuned DiT https://aiweekly.co/alerts/intel-passes-one-million-high-na-euv-wafers-ahead-of-rivals AI Weekly en 2026-09-09T09:28:43+00:00 Intel Passes One Million High-NA EUV Wafers, Ahead of Rivals https://aiweekly.co/alerts/openai-gpt-56-sol-ran-40-quantum-measurements-at-mit-lab AI Weekly en 2026-09-09T09:28:43+00:00 OpenAI: GPT-5.6 Sol Ran 40 Quantum Measurements at MIT Lab https://aiweekly.co/alerts/openai-agent-breach-fuels-push-for-ntsb-style-ai-probe AI Weekly en 2026-09-09T09:12:01+00:00 OpenAI agent breach fuels push for NTSB-style AI probe https://aiweekly.co/alerts/cohere-open-sources-decode-megakernel-158-vllm-on-one-h100 AI Weekly en 2026-09-09T09:12:01+00:00 Cohere open-sources decode megakernel, 1.58× vLLM on one H100 https://aiweekly.co/alerts/berkeleys-tango-ships-29-dof-humanoid-vla-on-unitree-g1 AI Weekly en 2026-09-09T07:56:04+00:00 Berkeley's TANGO Ships 29-DoF Humanoid VLA on Unitree G1 https://aiweekly.co/alerts/nex-agi-ships-35b-nex-n25-mini-agent-model-under-apache-20 AI Weekly en 2026-09-09T07:56:04+00:00 Nex-AGI Ships 35B Nex-N2.5-mini Agent Model Under Apache 2.0 https://aiweekly.co/alerts/oprd-paper-lets-student-llms-surpass-weak-teachers-faster AI Weekly en 2026-09-09T07:56:04+00:00 OPRD Paper Lets Student LLMs Surpass Weak Teachers Faster https://aiweekly.co/alerts/cover-retains-935-of-3d-vlm-accuracy-using-8-of-tokens AI Weekly en 2026-09-09T07:56:04+00:00 CoVeR Retains 93.5% of 3D VLM Accuracy Using 8% of Tokens https://aiweekly.co/alerts/anthropic-skips-uk-aisi-pre-release-testing-of-mythos-51 AI Weekly en 2026-09-09T06:38:34+00:00 Anthropic Skips UK AISI Pre-Release Testing of Mythos 5.1 https://aiweekly.co/alerts/chinas-csrc-raises-humanoid-ipo-bar-after-unitree-selloff AI Weekly en 2026-09-09T06:38:34+00:00 China's CSRC Raises Humanoid IPO Bar After Unitree Selloff https://aiweekly.co/alerts/auk-ships-15b-open-speech-model-unifying-generation-editing AI Weekly en 2026-09-09T06:38:34+00:00 AuK Ships 1.5B Open Speech Model Unifying Generation, Editing https://aiweekly.co/alerts/hugging-faces-mitchell-co-signs-call-to-drop-agi-north-star AI Weekly en 2026-09-09T06:12:01+00:00 Hugging Face's Mitchell co-signs call to drop AGI north star https://aiweekly.co/alerts/stella-parton-asks-internet-to-stop-ai-tributes-to-dolly AI Weekly en 2026-09-09T06:12:01+00:00 Stella Parton asks internet to stop AI tributes to Dolly https://aiweekly.co/alerts/anthropic-alignment-lead-puts-ai-extinction-odds-over-10 AI Weekly en 2026-09-09T04:58:06+00:00 Anthropic Alignment Lead Puts AI Extinction Odds Over 10% https://aiweekly.co/alerts/cmu-mole-benchmark-72-of-ai-agents-finish-insider-attacks AI Weekly en 2026-09-09T03:54:24+00:00 CMU MOLE benchmark: 72% of AI agents finish insider attacks https://aiweekly.co/alerts/live-llm-trading-fleet-posts-41-win-rate-vs-retails-50 AI Weekly en 2026-09-09T03:54:24+00:00 Live LLM Trading Fleet Posts 41% Win Rate vs Retail's 50% https://aiweekly.co/alerts/drivezero-beats-nuplans-log-replay-expert-without-human-demos AI Weekly en 2026-09-09T03:54:24+00:00 DriveZero beats nuPlan's log-replay expert without human demos https://aiweekly.co/alerts/dwarkesh-study-data-beat-model-recipes-324x-in-pretraining AI Weekly en 2026-09-09T03:36:23+00:00 Dwarkesh Study: Data Beat Model Recipes 3.24x in Pretraining https://aiweekly.co/alerts/neohorse-1-pitches-routing-loop-as-recursive-self-improvement AI Weekly en 2026-09-09T03:36:23+00:00 NeoHorse-1 Pitches Routing Loop as Recursive Self-Improvement https://aiweekly.co/alerts/alpha-school-wires-teens-to-heart-monitors-for-critique-videos AI Weekly en 2026-09-09T03:12:01+00:00 Alpha School wires teens to heart monitors for critique videos https://aiweekly.co/alerts/krugs-astra-plays-montezumas-revenge-an-old-agi-marker AI Weekly en 2026-09-09T03:12:01+00:00 Krug's Astra plays Montezuma's Revenge, an old AGI marker https://aiweekly.co/alerts/openai-claims-navierstokes-proof-nyu-pair-disputes-credit AI Weekly en 2026-09-09T03:12:01+00:00 OpenAI claims Navier–Stokes proof; NYU pair disputes credit https://aiweekly.co/alerts/pi-extension-makes-ai-agents-write-arm64-assembly-for-files AI Weekly en 2026-09-09T03:12:01+00:00 Pi extension makes AI agents write ARM64 assembly for files https://aiweekly.co/alerts/runway-ships-free-adobe-premiere-and-after-effects-plugins AI Weekly en 2026-09-09T03:12:01+00:00 Runway Ships Free Adobe Premiere and After Effects Plugins https://aiweekly.co/alerts/openai-claims-navier-stokes-proof-as-rival-teams-post-their-own AI Weekly en 2026-09-09T03:12:01+00:00 OpenAI claims Navier-Stokes proof as rival teams post their own https://aiweekly.co/alerts/anthropic-researcher-coxon-resigns-warns-of-ai-endgame AI Weekly en 2026-09-09T01:57:18+00:00 Anthropic Researcher Coxon Resigns, Warns of AI 'Endgame' https://aiweekly.co/alerts/i-have-adhd-skill-trends-on-hn-cuts-coding-agent-preamble AI Weekly en 2026-09-09T01:57:18+00:00 'i-have-adhd' Skill Trends on HN, Cuts Coding-Agent Preamble https://aiweekly.co/alerts/anthropic-sued-over-claude-max-5x-and-20x-usage-claims AI Weekly en 2026-09-09T00:31:29+00:00 Anthropic Sued Over Claude Max '5x' and '20x' Usage Claims https://aiweekly.co/alerts/intercept-foia-reveals-200m-pentagon-ai-deals-with-four-labs AI Weekly en 2026-09-08T21:28:15+00:00 Intercept FOIA Reveals $200M Pentagon AI Deals With Four Labs https://aiweekly.co/alerts/anthropic-quits-iti-after-push-to-strip-chip-export-bills AI Weekly en 2026-09-08T21:28:15+00:00 Anthropic Quits ITI After Push to Strip Chip Export Bills https://aiweekly.co/alerts/openais-navier-stokes-claim-rocked-by-nyu-misconduct-charge AI Weekly en 2026-09-08T21:12:01+00:00 OpenAI's Navier-Stokes claim rocked by NYU misconduct charge https://aiweekly.co/alerts/anthropic-linear-probes-catch-sleeper-agents-at-99-auroc AI Weekly en 2026-09-08T21:12:01+00:00 Anthropic linear probes catch sleeper agents at 99% AUROC https://aiweekly.co/alerts/ace-halves-moe-expert-use-without-training-or-calibration AI Weekly en 2026-09-08T21:12:01+00:00 ACE halves MoE expert use without training or calibration https://aiweekly.co/alerts/anthropic-faces-class-action-over-claude-max-usage-limits AI Weekly en 2026-09-08T21:12:01+00:00 Anthropic faces class action over Claude Max usage limits https://aiweekly.co/alerts/openai-mathematician-clash-over-ais-navier-stokes-proof-claim AI Weekly en 2026-09-08T21:12:01+00:00 OpenAI, mathematician clash over AI's Navier-Stokes proof claim https://aiweekly.co/alerts/meta-debuts-muse-ai-agent-that-runs-in-its-own-secure-vm AI Weekly en 2026-09-08T19:58:14+00:00 Meta Debuts Muse AI Agent That Runs in Its Own Secure VM https://aiweekly.co/alerts/openai-adds-sketch-to-chatgpt-images-25-ships-two-api-models AI Weekly en 2026-09-08T19:58:14+00:00 OpenAI Adds Sketch to ChatGPT Images 2.5, Ships Two API Models https://aiweekly.co/alerts/nandy-commits-to-law-forcing-apple-google-to-block-kids-nudes AI Weekly en 2026-09-08T18:29:28+00:00 Nandy Commits to Law Forcing Apple, Google to Block Kids' Nudes https://aiweekly.co/alerts/tomasello-uniquely-human-cognition-rooted-in-shared-attention AI Weekly en 2026-09-08T18:12:02+00:00 Tomasello: uniquely human cognition rooted in shared attention https://aiweekly.co/alerts/deepmind-ships-alphagenome-atlas-9b-dna-variant-predictions AI Weekly en 2026-09-08T18:12:02+00:00 DeepMind ships AlphaGenome Atlas: 9B DNA variant predictions https://aiweekly.co/alerts/open-model-licenses-split-as-chinese-labs-add-revenue-gates AI Weekly en 2026-09-08T18:12:02+00:00 Open-model licenses split as Chinese labs add revenue gates https://aiweekly.co/alerts/mistral-posts-nvfp4-build-of-small-4-119b-with-vllm-red-hat AI Weekly en 2026-09-08T18:12:02+00:00 Mistral posts NVFP4 build of Small 4 119B with vLLM, Red Hat https://aiweekly.co/alerts/saphra-and-wiegreffe-map-four-uses-of-mechanistic-in-ai AI Weekly en 2026-09-08T18:12:02+00:00 Saphra and Wiegreffe map four uses of 'mechanistic' in AI https://aiweekly.co/alerts/openai-claims-navier-stokes-proof-buckmaster-alleges-misconduct AI Weekly en 2026-09-08T16:58:21+00:00 OpenAI Claims Navier-Stokes Proof; Buckmaster Alleges Misconduct https://aiweekly.co/alerts/nextera-wins-19b-doe-loan-to-restart-iowa-reactor-for-google AI Weekly en 2026-09-08T16:58:21+00:00 NextEra Wins $1.9B DOE Loan to Restart Iowa Reactor for Google